impose_frame_attributes#

astropy.coordinates.impose_frame_attributes(**kwargs)[source]#

Assume a value of one or more frame attributes.

Parameters:

kwargs: This function accepts any keyword arguments and will override any frame attribute with a name matching that of the keyword name.

Examples

>>> import astropy.units as u
>>> from astropy.coordinates import EarthLocation, SkyCoord
>>> from astropy.coordinates.attributes import impose_frame_attributes

>>> location = EarthLocation(-5466045*u.m, -2404388*u.m, 2242133*u.m)

Say you have two coordinates that represent points in images which are very close in time but not exactly the same:

>>> coord1 = SkyCoord(10, 20, unit=u.deg, obstime="2026-01-01T00:00:00", location=location, frame="altaz")
>>> coord2 = SkyCoord(10, 20, unit=u.deg, obstime="2026-01-01T00:01:00", location=location, frame="altaz")

If you transform these to other frames, they will give slightly different results:

>>> coord1.transform_to("icrs")
<SkyCoord (ICRS): (ra, dec) in deg
    (36.10412254, 80.47633131)>
>>> coord2.transform_to("icrs")
<SkyCoord (ICRS): (ra, dec) in deg
    (36.35158456, 80.47671193)>

If however, you are happy to ignore this small change in observer location, and you want to avoid more complex transformations or other effects through the transform graph, you can do this:

>>> with impose_frame_attributes(obstime="2026-01-01T00:00:00"):
...     coord2.transform_to("icrs")
<SkyCoord (ICRS): (ra, dec) in deg
    (36.10412254, 80.47633131)>

Another example is computing the separation between two HCRS frames with slightly different obstimes.

>>> import astropy.units as u
>>> from astropy.coordinates import SkyCoord

>>> coord1 = SkyCoord(10*u.deg, 20*u.deg, frame='hcrs', obstime='2026-01-01 00:00:00')
>>> coord2 = SkyCoord(20*u.deg, 30*u.deg, frame='hcrs', obstime='2026-01-01 00:00:00.001')

When you try and compute the separation an error is raised because of the origin shift.

>>> print(coord1.separation(coord2))
...
astropy.units.errors.UnitsError: The input HCRS coordinates do not have length units. This probably means you created coordinates with lat/lon but no distance.  Heliocentric<->ICRS transforms cannot function in this case because there is an origin shift.

However, using this context manager:

>>> with impose_frame_attributes(obstime=coord1.obstime):
...     print(coord1.separation(coord2))
13d28m54.20360928s

Both of these examples are trivial when constructing coordinates like this, but when operating on a large number of coordinates read from a file or with coordinate transforms in packages such as reproject, it becomes harder to make these changes manually.